Clusters composed of the following variables: Overall mortality due to NCDs (Non-Communicable Diseases) adjusted for age for the period from 2010 to 2019. Overall mortality due to communicable diseases adjusted for age for the period from 2010 to 2019. Overall mortality due to external causes adjusted for age for the period from 2010 to 2019. Overall mortality due to ill-defined causes adjusted for age for the period from 2010 to 2019. Overall mortality due to other causes adjusted for age for the period from 2010 to 2019. Maternal mortality of women of reproductive age for the period from 2010 to 2019. Perinatal mortality for the period from 2010 to 2019. Hospital morbidity in the SUS (Brazilian Unified Health System) due to NCDs adjusted for age for the period from 2010 to 2019. Hospital morbidity in the SUS due to communicable diseases adjusted for age for the period from 2010 to 2019. Hospital morbidity in the SUS due to external causes adjusted for age for the period from 2010 to 2019. Hospital morbidity in the SUS due to ill-defined causes adjusted for age for the period from 2010 to 2019. Hospital morbidity in the SUS due to other causes adjusted for age for the period from 2010 to 2019. Hospital morbidity in the SUS due to maternal causes of women of reproductive age for the period from 2010 to 2019. Hospital morbidity in the SUS due to perinatal causes for the period from 2010 to 2019.
* The mortality coefficients are based on 100,000 inhabitants.
* The hospitalar morbidity coefficients are based on 10,000 SUS-dependent inhabitants.
* All coefficients were normalized using z-scores for the formation of the clusters.
* The SKATER algorithm was used for cluster analysis, as the goal was to create contiguous regions with similar morbidity and mortality profiles.
* The queen method of order 1 was adopted as the neighborhood criterion. The municipality of Ilhabela, which does not have direct neighbors, was imputed to the nearest cluster.